Model Based Diagnostics of Servo-Hydraulic System Using Quasi-Adaptive System Identification Approach

The aim of this paper is to provide a feasibility study of model based diagnostics using a "black-box" model approach. A moderately complex first principle model of a servo-hydraulic system was proposed to understand the diagnostic symptoms. Performance of a steam turbine control system is one of key factors for proper management of power plant assets. To achieve quality goals more advanced approach to diagnostic needs to be implemented. This paper presents model-based approach to diagnostics of hydraulic actuator performance monitoring eignefrequency and damping ratio variability. The feasibility study presented in this paper confirmed correctness of the proposed approach using quasi-adaptive system identification technique involving an ARX model. The results can also be utilized for optimal adjustment of the controller settings considering the load and variable stiffness of the actuator dependent on the piston-rod position.